Minutes - CA-forum #15
Friday 24 May, 1996; 11:00 to 12:30
venues : epping : downstairs conference room
narrabri :
1. Precision of stored positions (caobs/sched) - (RS, DMcK)
The problems identified by J.Reynolds last november have been rectified.
The limitation (100 msec in RA, 10 mas in DEC)
imposed by the caobs-cacor message format seems
unecessary, and can be remedied: Caobs should send double precision,
rather than an ascii string.
>> WEW and the Narrabri group will make the change.
2. CACAL and the tied array - (DMcC)
WEW recommended that this upgrade should be put on hold until the
new Tied-Array hardware is installed. (Expected in the next month or so).
3. Antenna 4 - narrow band poln. leakage. - (WEW/BR)
WEW and BR found a problem with the XY correlator of antenna 4. (A cable
interchange). The problem identified during the tied-array tests in
april has been eliminated, and antenna #4 can now be used as reference
antenna.
The problem identified by RS, (the frequency-dependant polarisation
leakage problem) once thought to have the same origin is
unlikely to be related to the cabling interchange. Bob's recent
tests (may 22) show that the problem is still present, although
now somewhat weaker. Bob will repeat the experiment.
4. DUTC updating from the correlator (WEW, DMcC)
WEW suggests that the correlator, which can read the Hunt-clock timeframe
should take on the responsability of distributing the Delta_UTC (number of
leap seconds) to the wider network.
>> WEW and Narrabri will look into this.
5. Reference Pointing - (DMcK)
(see also : http://wwwnar/www/operations/ref_pointing.html)
This is now available on a routine basis. Everyone, Especially
local experts, is encouraged to make use of it.
The note on the web provides an outline of what the observer needs to do,
and what can be expected.
There are number of issues which will need to be resolved - eg, how does
one "age" the reference solution, and should one warn the observer. These
questions are best tackled in the light of observers using the system.
Bob Sault raised the question of the pointing error algorithm in the face
of the observed C and X band beam asymetries and squints.
>> Bob will look into this.
6. Sac Switch problems - (DMcK, NebK)
This relates to the handling of the setting of attenuators and sampler levels
during mosaicing. For almost all applications the machinery works adequately.
However, one experiment (CenA), is technically challenging, and poses
difficulties. The SAC machinery was revised for this. The machinery
is fragile in this context.
a. If CEN A is the only experiment of this kind, then no further action
is required -- N.Killeen can improvise.
b. Users who are concerned that they also will experience trouble
(ie, if the Tsys level varies considerably from field to field) should
consult N.Killeen or D.McConnell.
c. If there is a real need, then substantial effort will have to be
expended, as the attenuator setting will need to be shifted to the ACC.
7. Data Quality Observations - (JH/RW)
- the analysis is now up to date. JH is revising the output format -
with plots and tables that provide a better view of the array status.
These will be available on the web
(http://wwwnar.atnf.csiro.au/www/operations/dqo/dqo.html and references
thereafter).
- Jim Higdon is processing the blank field now that a dozen or so
images have been observed, in order to obtain the deep averaged field.
This will provide an indication of what will be achievable for comparison
with the Hubble blank field experiment.
- The HST have a short list of 8 Blank field candidates,
shortly to be pruned to perhaps
four. We will observe these candidates, and possibly provide guidance in the
selection of the final field, which we would propose to make our standard
for the DQO. (J.Higdon and colleagues from narrabri, R.Norris and R.Sault).
8. Remote focus matters - (DMcK)
The hardware is in place - the subreflector can be driven from the
control room, under manual control. Computer control is not yet
available.
- R.Beresford's advice is solicited to determine to remaining difficulties
in going to computer control -- is it just a matter of calibration?
- M.Kesteven will provide the focus curves, so that the cost (to
observers) of not focussing can be assessed.
9. Web page - "New features/problems"
A request has been received for a web page which lists (briefly) new
information which may affect observers - new utilities (eg, reference
pointing), bugs (eg, the incorrect handling of non-J2000 epochs).
M.Kesteven will look after this.
Follow-ups:
S-band poln - see Bob's reports in his home page :
http://www.atnf.csiro.au/~rsault
The ATCA 13cm Polarimetric Response (Sault and Ehle), and
ATCA Off-axis Polarimetric Response (Sault).
Tied-array - see note 2
----------------------------- Notes --------------------------------------
1. Stored position precision (From R.Sault, D.McKay)
Derek and I are happy that coordinate "precision" is not being lost in
SCHED/CAOBS (apart from rounding when loaded into CAOBS to the nearest
millisec of RA and 10 milliarcsec in DEC -- because of the CAOBS/CACOR
communication limitation). Unless there are more sublte bugs (the coordinates
for 0823 and 1934 are getting through OK), the problems that John raised
last year were fixed in November. The problem that Derek/Dave noted last
CA-forum are traceable back to a SCHED file that was almost certainly
produced before November.
We probably have to address whether the CAOBS/CACOR limitation
is worth fixing.
2. Split Array
The position appears to be this:
- control of the array (caobs) is probably able to manage several
sub-arrays... it was so tested several years ago. I did hear that
there were LO conflicts, but I doubt that they are fundamental.
- it is very likely that all the sub-arrays must share the
same intergation cycling.
- VLBI now uses the delay units to convert from sampled data
to S2 format. At present the delay unit does not support split
functions. It is just conceivable that as long as both sub-arrays
are in the same mode (4-bit, narrow band) coexistance is
possible, but clearly unsatisfactory.
- Warwick estimates that it will require several months of effort
from an Evan Davis-clone (ie, Warwick, at this stage) in order to
rework the delay unit.
------------------------------------------------------------------
http://www.atnf.csiro.au/~mkesteve/ca-forum.html